Method of and apparatus for forming halftone dots

Abstract

 Halftone dots are formed on the basis of three sets of screen pattern data assinged to a same combined area, in which a plurality of halftone dots are to be formed. The three sets of screen pattern data have different threshold values assinged to respective pixels in the combined area, and are stored in SPM's (101-116): eight SPM's store a first set of the screen pattern data, six SPM's store a second one, and two SPM's store a third one. One of the sixteen SPM's is randomly seleoted with respeot to each pixel, whereby one of the threshold values (D_s) stored in the selected SPM is compared with an image singal (V₁). A dot signal indicating whether or not each pixel is solid is produced on the basis of the comparison. Accordingly, a grid-like pattern of halftone dots to be seen in a highlight area or a shadow area is prevented.

Description

BACKGROUND OF THE INVENTION

Field of the Invention

[0001] The present invention is related to a method of and an apparatus for forming halftone dots in a halftone image, which is recorded on photosensitive film or the like

Description of the Prior Art

[0002] In general, halftone dots in a halftone image are formed as a result of comparison between image data and threshold data therefor, which is called screen pattern data. The image data are compared with the screen pattern data which respect to each pixel in an image represented by the image data, whereby a dot signal indicating whether or not each pixel is to be solid is produced. The halftone dots are formed of solid pixels which are exposed as a function of the dot signal.

[0003] The screen pattern data are so prepared that the size of a halftone dot increases on a positive halftone image as the density of the original image increases. Therefore, one set of screen pattern data are usually assinged to a full area of a halftone dot at the halftone-­dot area rate of 100% (hereinafter referred to as "full halftone-dot area"). If each of the screen pattern data assigned to each pixel in the full halftone-dot area has a different value from each other, the number of tone gradation levels represented by the screen pattern data is equal to the number of pixels in the full halftone-dot area.

[0004] On the other hand, some technique, such as one disclosed in Japanese Patent Laying Open Gazette No. 63-­97067, is proposed in order to attain a greater number of tone gradation levels while maintaining the size of the full halftone-dot area, that is, while maintaining a screen ruling. This technique employs a combined area which consists of a plurality of full halftone-dot areas, and combined screen pattern data whose values are different from each other and which are assigned to the pixels in the combined area. For example, if the combined area consists of two full halftone-dot areas each of which is a 8×8 pixel matrix, the combined screen pattern data represent 128 levels of tone gradation.

[0005] However, a halftone image produced with the combined screen pattern data may cause the following problem. In a part of a halftone image where density is uniform, a fixed shape of halftone dots is repeated because the combined screen pattern data are assigned repeatedly on an image. Figs. 1A and 1C illustrate enlarged views of a highlight area and a shadow area thus formed, respectively. The highlight area usually has the halftone-dot area rate of at most 10%, and the shadow area usually has a halftone-dot area rate of at least 90%. The highlight area of Fig. 1A includes a first solid portion consisting of one solid dot and a second solid portion consisting of two solid dots. The shadow area in Fig. 1C similarly includes a few types of blank portions. The solid portions and the blank portions are formed repeatedly in respective areas. Consequently, the array of the solid portions and that of the blank portions in Figs. 1A and 1C can be seen as grid-like patterns as shown in Figs. 1B and 1D, respectively, with the naked eye. Since these grid-like patterns do not exist in an original image, they decrease the quality of the halftone image. This problem is important especially in the highlight area and the shadow area.

SUMMARY OF THE INVENTION

[0006] The present invention is directed to a method and an apparatus therefore for forming halftone dots in a halftone image, comprising the steps of: (a) preparing image data representing an original image, (b) preparing plural sets of threshold data, each of which includes predetermined threshold values assigned to respective pixels in a unit pixel block of a prescribed shape, said plural sets of threshold data having different arrangement of the threshold values from each other, (c) fixing the correspondance of each pixel in the original image to each pixel in the unit pixel block, (d) randomly selecting one of the plural sets of threshold data with respect to each pixel to thereby pick up one of the threshold values in a selected set of threshold data at each pixel, (e) comparing the image data with a threshold value picked up at the step (d) with respect to each pixel, to thereby produce a dot signal indicating whether or not each pixel is to be solid, and (f) forming halftone dots as a function of the dot signal. Acc­ording to an spect of the present invention, each of the plural sets of threshold data is so established that a plurality of halftone dots are to be formed in the unit pixel block.

[0007] According to another aspect of the present invention, the selecting at the step (d) is executed while probabilities of selecting the respective plural sets of threshold data are fixed in advance.

[0008] According to still another aspect. of the present invention, the threshold values in each of the plural sets of threshold data are different from each other.

[0009] According to another aspect of the present invention, the plural sets of threshold data are so established that the halftone dots are formed checkerwise in the unit pixel block when the image data for the unit pixel block has a data value corresponding to a halftone-dot area rate of at most about 10 percent.

[0010] According to still another aspect of the present invention, the plural sets of threshold data are so established that blank areas at which the halftone dots do not exist are formed checkerwise in the unit pixel block when the image data for the unit pixel block has a data value corresponding to a halftone-dot area rate of at least about 90 percent.

[0011] Preferably, differences between the threshold values of the plural sets of threshold data at each pixel in the unit pixel block are within a prescribed range.

[0012] The threshold values at each pixel in the unit pixel block may be common to the plural sets of threshold data in a range of the threshold values corresponding to a halftone-­dot area rate ranging from at least about 10 percent to at most about 90 percent.

[0013] Accordingly, an object of the present invention is to prevent grid-like patterns or line patterns in the highlight and shadow areas of a halftone image.

[0014] These and other objects, features, aspects and advantages of the present invention will become more apparent from the following detailed description of the present invention when taken in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

[0015] 

Figs. 1A through 1D illustrate grid-like pattern in the prior art;

Fig. 2 is a diagram showing the structure of an image scanner according to a preferred embodiment of the present invention;

Figs. 3A through 3C illustrate the arrangement of three sets of screen pattern data;

Figs. 4A through 4C illustrate grouping of unit areas in a combined area;

Figs. 5 and 7 are block diagrams showing the internal structure of a screen pattern memory unit; and

Fig. 6 illustrates halftone dots according to the preferred embodiment.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] Fig. 2 is a diagram showing the structure of a monochrome scanner of a flat bed type according to a preferred embodiment of the present invention. This color image scanner comprises a reading scanner unit 1 and a recording scanner unit 2.

[0017] An original 4 is mounted on a transparent plate 3 in the reading scanner unit 1. A flourescent lamp 6 covered with a reflecting plate 5 is installed under the transparent plate 3, and light from the fluorescent lamp 6 is incident and reflected on the original 4. The light representing an image of the original 4 is reflected by a mirror 7, and passing through a lens 8 to converge on a CCD 9. The image of the original received by the CCD 9 is contracted by the lens 8. Beside the transparent plate 3 is provided a white reference plate 10 with which several characteristics of the original image, such as illuminance distribution on the original 4, cos⁴ϑ characteristics, mismatch among internal devices of the CCD 9 and the like, are compensated. Main scanning is acheived by relatively moving the plate 3 and the optical system by means of drive mechanism (not shown) in an x-direction at a constant speed corresponding to a prescribed magnification factor in the reading operation.

[0018] An image processing unit 11 receives an image signal V₀ and executes image processing, such as gradation correction, sharpness intensification, magnification and the like, thereby producing a processed image signal V₁. The processed image signal V₁ is supplied from the image processing unit to a SPM unit (Screen Pattern Memory unit) 12. The SPM unit 12 comprises several SPM's for storing threshold data for the processed image signal V₁ as described later in detail. The threshold data will be hereinafter referred to as "screen pattern data". The processed image signal V₁ is compared with the threshold data in the SPM unit 12 in synchronism with a start signal S_s supplied from a start sensor 14 provided in the recording unit 2, whereby a dot signal S_d is produced. Exposure light in the recording scanner unit 2 is ON/OFF controlled by the dot signal S_d.

[0019] A film-feeding roller 15 is rotated by a subscanning drive motor 16, thereby feeding a film 17 as photosensitive material for recording images in a subscanning direction X.

[0020] A semiconductor laser 18 provided in the recording scanner unit 2 generates a laser beam modulated on the basis of the dot signal S_d supplied from the SPM unit 12. The laser beam, which is initially diverging, is made into a parallel beam by a collimate lens 19, passing through a cylindrical lens 20 to converge on a mirror surface of a polygon mirror 21.

[0021] The polygon mirror 21 comprises a hexagonal mirror for reflecting and deflecting the laser beam. Each mirror surface of the hexagonal mirror deflects the laser beam to perform scanning on one main scanning line. The laser beam reflected and deflected by the polygon mirror 21 passes through an fϑ lens 22 and a cylindrical lens 23, and finally scans the surface of the film 17. The fϑ lens 22 operates so that a beam spot of the laser beam on the film 17 has a constant size in any position on the main scanning line, and that the beam spot moves in proportion to an angle of rotation of the polygon mirror 21, that is, at a constant speed. The cylindrical lens 23 corrects the laser beam to mainly compensate fabrication errors of the polygon mirror 21.

[0022] Above the film 17, a mirror 24 and the start sensor 14 comprising photosensitive device such as photodiode are installed to detect the laser beam to thereby detect the beginning of each main scanning operation. The start sensor 14 supplies the start signal S_s to the SPM unit 12 as described before to control timing for producing the dot singal S_d.

[0023] Figs. 3A through 3C illustrate the arrangement of three sets of screen pattern data. The following explanation will be made about positive halftone image where a highlight area is formed of small halftone dots, for convinience of illustration.

[0024] Each of the three sets of screen pattern data P_A, P_B and P_C shown in Figs. 3A through 3C, respectively, is assigned in a combined area (or a unit pixel block) A_c. The combined area A_c includes sixteen unit areas M. One halftone dot is formed in a pair of unit areas as described below. The unit area M, which is shaded in Fig. 3A, is a 5×5 matrix of elementary areas (or pixels) K. The screen pattern data are compared with the processed image signal V₁, whereby the dot signal S_d for controlling the exposure light is produced as described before. Therefore, the arrangement of the screen pattern data are established so that halftone-dot area rate in the combined area A_c become proportional to the level of the processed image signal V₁. Figs. 3A through 3C show only a part of the values of the screen pattern data. Since the combined area A_c consists of 400 elementary areas K, the screen pattern data can have as many as 400 gradation levels if the elementary areas have respective values different from each other. The screen pattern data PA shown in Fig. 3A has basic data arrangement, and the data arrangement of the other screen pattern data PB and PC shown in Figs. 3B and 3C, respectively, are slightly different from the basic data arrangement.

[0025] Fig. 4 also illustrates the structure of the combined area Ac. The combined area is a 4×4 matrix consisting of unit areas M₁₁-M₄₄. Shaded unit areas in Fig. 4B are to be exposed at least partly at the halftone-dot area rate of 50 percent or less, and shaded unit areas in Fig. 4C are to be exposed at least partly in addition to the shaded unit areas of Fig. 4B at the halftone-dot area rate of 51 percent or more. Namely, halftone dots (or solid portions) are formed checkerwise in the unit areas M₁₁, M₁₃, M₂₂, M₂₄, M₃₁, M₃₃, M₄₂ and M₄₄ at the halftone dot area rate of 50 percent or less, and are formed additionally in the unit areas M₁₂, M₁₄, M₂₁, M₂₃, M₃₂, M₃₄, M₄₁ and M₄₃ at the halftone-dot area rate of 51 percent or more. Therefore, blank areas where halftone dots do not exist are formed checkerwise in the combined area when the combined area is in a shadow area.

[0026] The unit areas M₁₁-M₄₄ are classified into four groups G₁-G₄. The first group G₁ consists of the unit areas M₁₁, M₁₃, M₃₁ and M₃₃ as shown in Fig. 4B, and the second group G₂ consists of the unit areas M₂₂, M₂₄, M₄₂ and M₄₄. Similarly, as shown in Fig. 4C, the third group G₃ consists of the unit areas M₁₂, M₁₄, M₃₄ and M₃₄ and the fourth group G₄ consists of the unit areas M₂₁, M₂₃, M₄₁ and M₄₃. As can be seen from Figs. 3A and 4A, the values from 1 to 4 of the screen pattern data P_A are assigned to the unit areas in the group G₁, the values from 5 to 8 in the group G₂, the values from 9 to 12 in the group G₁, the values from 13 to 16 in the group G₂. In other words, four sequential values of the screen pattern data are assigned alternately to the groups G₁ and G₂ at the halftone-dot area rate of 50 percent or less. Similarly, four sequential values of the screen pattern data are assigned alternately to the groups G₃ and G₄ at the halftone-dot area rate of 51 percent or more. This is the same with the other screen pattern data P_B and P_C. The three sets of screen pattern data P_A, P_B and P_C are different only in the order of assignment of the four sequential values in each group: for example, the values of 1 to 4 are respectively assigned to the unit areas M₁₁, M₃₃, M₁₃ and M₃₁ in this order in the screen pattern data P_A, to M₃₃, M₁₁, M₃₁ and M₁₃ in the screen pattern data P_B, and to M₃₁, M₁₃, M₃₃ and M₁₁ in the screen pattern data P_C. Consequently, all of the screen pattern data P_A, P_B and P_C give the same number of solid pixels in the combined unit area A_c with respect to a given processed image data. In other words, the three screen pattern data give the same halftone-dot area rate representing density of an original image, and represent different pattern of solid pixels from each other. Further, as a result of the way of the assignment of the four sequential values, the differences between the values of the three screen pattern data P_A, P_B and P_C are within plus or minus three at a same position in the combined area A_c. These differences are so determined that the reproduced image represented by the halftone dots can attain similar tone gradation to an original image. In other words, if the differences are too large, the tone gradation of the reproduced image becomes much different from that of the original image.

[0027] Fig. 5 is a block diagram .showing the internal structure of the SPM unit 12. The SPM unit 12 comprises sixteen SPM's 101-116. Out of the sixteen SPM's, eight SPM's store the basic screen pattern data P_A, six SPM's store the screen pattern data P_B, and two SPM's store the screen pattern data P_C. Since one of the sixteen SPM's is randomly selected with respect to each pixel, as described later, the numbers of the SPM's for respective screen pattern data mean that the three screen pattern data P_A, P_B and P_C have substantial weighting factors of four, three and one, respectively.

[0028] The processed image signal V₁ is expressed by digital data of 12 bits, for example. The digital data of 12 bits can express four thousand gradation levels. However, because the screen pattern data P_A, P_B and P_C have the four hundred gradation levels in this embodiment, for convenience of illustration, the processed image signal V₁ only represent the four hundred gradation levels.

[0029] The processed image signal V₁ of 12 bits is supplied to a comparator 125, and a least-bit signal V_1a a consisting of the least four bits of the processed image signal V₁ is supplied to a SPM selector 123.

[0030] The least-bit signal V_1a is a noise component which is given to the image signal when it is obtained through reading an original. Therefore, the least-bit signal V_1a has a random value ranging from zero to sixteen. The SPM selector 123 decodes the least-bit signal V_1a to produce sixteen selection signals S₁-S₁₆, one of which has "H" level correspondingly to the value of the least bit signal V_1a. The selection signals S₁-S₁₆ are supplied to the SPM's 101-­116, respectively, thereby randomly selecting one of the SPM's which receives "H" level signal.

[0031] Incidentally, the selection signals may be produced in a random number generator well known in the art. In this case, the SPM unit 12 may include only three SPM's for storing the three screen pattern data P_A, P_B and P_C. The least-bit signal V_1a is converted into three selection signals which are weighted by the weighting factors in a signal conversion means.

[0032] The SPM unit 12 further comprises an address counter 124. The address counter 124 produces an address signal A_d to be given to the SPM's 101-116 on the basis of a positional signal (not shown) supplied from an encoder (not shown) which generates the positional signal representing the recording position on the film 17. The address signal A_d indicates one of the elementary area K in the combined area A_c. One SPM which is selected by the selection singal S₁-S₁₆ outputs a threshold value D_s for the elementary area (or pixel) indicated by the address signal A_d. The threshold data are supplied to the comparator 125 as well as the processed image signal V₁.

[0033] The comparator 125 compares the level of the processed image signal V₁ with the threshold data D_s with respect to each pixel in order of scanning positions, thereby producing the binary dot signal S_d. The dot signal Ss is supplied to the semiconductor laser 18, as described before, whereby the laser beam is ON/OFF controlled and a halftone image is recorded on the film 17.

[0034] Fig. 6 illustrates a highlight area of a halftone image recorded according to the preferred embodiment of the present invention. In Fig. 6, the level of the processed image signal is assumed to be 3.5. Solid lines drawn in Fig. 6 indicates the boundaries between the combined areas A_c. As shown in Fig. 6, each combined area has three or four halftone dots, and the positions of the three or four halftone dots in each combined areas are occasionally different from each other. Since the threshold value is selected randomly with respect to each elementary area (or each pixel), each combined area shown in Fig. 6 does not correspond to one of the three screen pattern data P_A, P_B and P_C. However, since the positions of the halftone dots in the combined areas are occasionally different from each other, these halftone dots cannot be seen as a grid-like lines as shown in Figs. 1A through 1D.

[0035] In addition, the halftone dots shown in Fig. 6 have almost same size, even if they are placed at various positions, so that the tone gradation of the original image is faithfully represented by the halftone image; this is acheived by the fact that the differences between the threshold values at a same pixel in the three screen pattern data are within plus or minus three.

[0036] Incidentally, the screen pattern data P_A, P_B and P_C are not necessarily weighted by weighting factors. Namely, the SPM unit 12 may includes only three SPM's for storing the three screen pattern data, respectively. However, by weighting the screen pattern data with respective weighting factors, it can be prevented that the halftone dots are placed collectively at some points and scarcely at other points, which is seen to include areas of deep density and light density.

[0037] The number of the screen pattern data is not limited to three, but it can be any more than two. Further, the weighting factos for respective screen pattern data can be changed arbitrarily.

[0038] The threshold values of the screen pattern data P_A, P_B and P_C for the shadow area, or an area having the halftone-dot area rate of about 90 percent or more, are arranged in the same manner as that for the highlight area described above. Further, the arrangement of the screen pattern data P_A, P_B and P_C can be same for middle tone areas, which have the halftone-dot area rate ranging from about 10 percent to about 90 percent, because the problem of the prior art is important only in the highlight area having the halftone-dot area rate of about 10 percent or less and in the shadow area having the halftone-dot area rate of about 90 percent or more. Fig. 7 is a block diagram showing the internal structure of the SPM unit 12 which employ an additional SPM 100 for storing the threshold values for the middle tone area. The SPM's 101a-116a store the threshold values only for the highlight and shadow areas. Since the total number of gradation levels required for the highlight and shadow areas are smaller than for the middle tone area, the threshold values stored in the SPM's 101a-116a can be expressed by digital data of less number of bits than those for the middle tone area stored in the SPM 100. Consequently, the threshold values are stored in the SPM's 101a-116a with less number of bits, which decreases the capacity and the cost of the SPM's 101a-116a.

[0039] The SPM unit 12 shown in Fig. 7 comprises a density modulation unit 122 which is connected to the SPM selector 123. The density modulation unit 122 comprises a gradation range judging unit 122a, a code setting unit 122b, and a ramdom number generation circuit 122c. The gradation range judging unit 122a receives the processed image signal V₁, and if the processed image signal V₁ represents a density level in the middle tone area, the gradation range judging unit 122a makes the code setting unit 122b output a prescribed code of 4-bit data indicating that the density level is in the middle tone area . If the processed image signal V₁ represents a density level out of the middle tone area, the gradation range judging unit 122a makes the ramdom number generation circuit 122c output a random number other than the prescribed code.

[0040] The SPM selector 123 selects the SPM 100 for the middle tone area if the prescibed code is supplied from the density modulation unit 122. Otherwise, the SPM selector 123 selects one of the SPM's 101a-116a in response to the random number supplied from the density modulation unit 122 in the same manner as the selector shown in Fig. 5.

[0041] Although the halftone image produced according to the above described embodiment is a positive image, a negative image can be made in a similar manner, where the descriptions about the highlight area and the shadow area are exchanged.

[0042] According to the present invention, the grid-like patterns and the line patterns in highlight and shadow areas of a halftone image can be prevented.

[0043] Although the present invention has been described and illustrated in detail, it is clearly understood that the same is by way of illustration and example only and is not to be taken by way of limitation. The spirit and scope of the present invention should be limited only by the terms of the appended claims.

Claims

1. A method of forming halftone dots in a halftone image, comprising the steps of:

(a) preparing image data representing an original image,

(b) preparing plural sets of threshold data, each of which includes predetermined threshold values assigned to respective pixels in a unit pixel block of a prescribed shape, said plural sets of threshold data having different arrangement of said threshold values from each other,

(c) fixing the correspondance of each pixel in said original image to each pixel in said unit pixel block,

(d) randomly selecting one of said plural sets of threshold data with respect to each pixel to thereby pick up one of said threshold values in a selected set of threshold data at each pixel,

(e) comparing said image data with a threshold value picked up at said step (d) with respect to each pixel, to thereby produce a dot signal indicating whether or not each pixel is to be solid, and

(f) forming halftone dots as a function of said dot signal.

2. A method in accordance with claim 1, wherein
each of said plural sets of threshold data is so established that a plurality of halftone dots are to be formed in said unit pixel block.

3. A method in accordance with claim 2, wherein
said selecting at said step (d) is executed while probabilities of selecting said respective plural sets of threshold data are fixed in advance.

4. A method in accordance with claim 3. wherein
said threshold values in each of said plural sets of threshold data are different from each other.

5. A method in accordance with claim 4, wherein
said plural sets of threshold data are so established that said halftone dots are formed checkerwise in said unit pixel block when said image data for said unit pixel block has a data value corresponding to a halftone-dot area rate of at most about 10 percent.

6. A method in accordance wiht claim 5, wherein
said plural sets of threshold data are so established that blank areas at which said halftone dots do not exist are formed checkerwise in said unit pixel block when said image data for said unit pixel block has a data value corresponding to a halftone-dot area rate of at least about 90 percent.

7. A method in accordance with claim 6, wherein
differences between said threshold values of said plural sets of threshold data at each pixel in said unit pixel block are within a prescribed range.

8. A method in accordance with claim 7, wherein
said threshold values at each pixel in said unit pixel block are common to said plural sets of threshold data in a range of said threshold values corresponding to a halftone-dot area rate ranging from at least about 10 peroent to at most about 90 percent.

9. An apparatus for forming halftone dots in a halftone image, comprising:

(a) means for providing image data representing an original image,

(b) means for storing plural sets of threshold data, each of which includes predetermined thereshold values assinged to respective pixels in a unit pixel block of a prescribed shape, said plural sets of threshold data having differenct arrangement of said threshold values from each other,

(c) means for fixing the correspondance of each pixel in said original image to each pixel in said unit pixel block,

(d) selecting means for randomly selecting one of said plural sets of threshold data with respect to each pixel to thereby pick up one of said threshold values in a selected set of threshold data at each pixel,

(e) means for comparing said image data with a threshold value picked up by said selecting means with respect to each pixel, and producing a dot signal indicating whether or not each pixel is to be solid, and

(f) means for forming halftone dots as a function of said dot signal.

10. An apparatus in accordance with claim 9, wherein
each of said plural sets of threshold data is so established that a plurality of halftone dots are to be formed in said unit pixel block.

11. An apparatus in accordance with claim 10, wherein
said selecting means selects one of said plural sets of threshold data while probabilities of selecting said respective plural sets of threshold data are fixed in advance.

12. An apparatus in accordance with claim 11, wherein
said threshold values in each of said plural sets of threshold data are different from each other.

13. An apparatus in accordance with claim 12, wherein
said plural sets of threshold data are so established that said halftone dots are formed checkerwise in said unit pixel block when said image data for said unit pixel block has a data value corresponding to a halftone-dot area rate of at most about 10 percent.

14. An apparatus in accordance with claim 13, wherein
said plural sets of threshold data are so established that blank areas at which said halftone dots do not exist are formed checkerwise in said unit pixel block when said image data for said unit pixel block has a data value corresponding to a halftone-dot area rate of at least about 90 percent.

15. An apparatus in accordance with claim 14, wherein
differences between said threshold values of said plural sets of threshold data at each pixel in said unit pixel block are within a prescribed range.

16. An apparatus in accordance with claim 15, wherein
said threshold values at each pixel in said unit pixel block are common to said plural sets of threshold data in a range of said threshold values corresponding to a halftone-dot area rate ranging from at least about precent to at most about 90 percent.

Drawing